It is small (16 B) and therefore the overhead of exporting it more
than outweighs the size savings from not having duplicated symbols:
When the symbol is no longer avpriv, one saves twice the size of
the string containing the symbols name (2x30 byte), two entries
in .dynsym (24 bytes each on x64), one entry in the importing libraries
.got and .rela.dyn (8 + 24 bytes on x64) and two entries for the
symbol version (2 bytes each) and one hash value in the exporting
library (4 bytes).
(The exact numbers are of course different for other platforms
(e.g. when using dlls), but given that the strings saved alone
more than outweigh the array size it can be presumed that this
is beneficial for all platforms.)
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
The child_class_next API relied on different AVCodecs to use
different AVClasses; yet this API has been replaced by
child_class_iterate.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@outlook.com>
Both AC-3 encoder share the same options, yet they are nevertheless
duplicated in the binary; and the options applying to the EAC-3 encoder
are a proper subset of the options for the AC-3 encoders, so that it can
use the same options as the former by putting the options specific to
AC-3 at the front. This commit implements this.
Signed-off-by: Andreas Rheinhardt <andreas.rheinhardt@gmail.com>
The AC3 encoder used to be a separate library called "Aften", which
got merged into libavcodec (literally, SVN commits and all).
The merge preserved as much features from the library as possible.
The code had two versions - a fixed point version and a floating
point version. FFmpeg had floating point DSP code used by other
codecs, the AC3 decoder including, so the floating-point DSP was
simply replaced with FFmpeg's own functions.
However, FFmpeg had no fixed-point audio code at that point. So
the encoder brought along its own fixed-point DSP functions,
including a fixed-point MDCT.
The fixed-point MDCT itself is trivially just a float MDCT with a
different type and each multiply being a fixed-point multiply.
So over time, it got refactored, and the FFT used for all other codecs
was templated.
Due to design decisions at the time, the fixed-point version of the
encoder operates at 16-bits of precision. Although convenient, this,
even at the time, was inadequate and inefficient. The encoder is noisy,
does not produce output comparable to the float encoder, and even
rings at higher frequencies due to the badly approximated winow function.
Enter MIPS (owned by Imagination Technologies at the time). They wanted
quick fixed-point decoding on their FPUless cores. So they contributed
patches to template the AC3 decoder so it had both a fixed-point
and a floating-point version. They also did the same for the AAC decoder.
They however, used 32-bit samples. Not 16-bits. And we did not have
32-bit fixed-point DSP functions, including an MDCT. But instead of
templating our MDCT to output 3 versions (float, 32-bit fixed and 16-bit fixed),
they simply copy-pasted their own MDCT into ours, and completely
ifdeffed our own MDCT code out if a 32-bit fixed point MDCT was selected.
This is also the status quo nowadays - 2 separate MDCTs, one which
produces floating point and 16-bit fixed point versions, and one
sort-of integrated which produces 32-bit MDCT.
MIPS weren't all that interested in encoding, so they left the encoder
as-is, and they didn't care much about the ifdeffery, mess or quality - it's
not their problem.
So the MDCT/FFT code has always been a thorn in anyone looking to clean up
code's eye.
Backstory over. Internally AC3 operates on 25-bit fixed-point coefficients.
So for the floating point version, the encoder simply runs the float MDCT,
and converts the resulting coefficients to 25-bit fixed-point, as AC3 is inherently
a fixed-point codec. For the fixed-point version, the input is 16-bit samples,
so to maximize precision the frame samples are analyzed and the highest set
bit is detected via ac3_max_msb_abs_int16(), and the coefficients are then
scaled up via ac3_lshift_int16(), so the input for the FFT is always at least 14 bits,
computed in normalize_samples(). After FFT, the coefficients are scaled up to 25 bits.
This patch simply changes the encoder to accept 32-bit samples, reusing
the already well-optimized 32-bit MDCT code, allowing us to clean up and drop
a large part of a very messy code of ours, as well as prepare for the future lavu/tx
conversion. The coefficients are simply scaled down to 25 bits during windowing,
skipping 2 separate scalings, as the hacks to extend precision are simply no longer
necessary. There's no point in running the MDCT always at 32 bits when you're
going to drop 6 bits off anyway, the headroom is plenty, and the MDCT rounds
properly.
This also makes the encoder even slightly more accurate over the float version,
as there's no coefficient conversion step necessary.
SIZE SAVINGS:
ARM32:
HARDCODED TABLES:
BASE - 10709590
DROP DSP - 10702872 - diff: -6.56KiB
DROP MDCT - 10667932 - diff: -34.12KiB - both: -40.68KiB
DROP FFT - 10336652 - diff: -323.52KiB - all: -364.20KiB
SOFTCODED TABLES:
BASE - 9685096
DROP DSP - 9678378 - diff: -6.56KiB
DROP MDCT - 9643466 - diff: -34.09KiB - both: -40.65KiB
DROP FFT - 9573918 - diff: -67.92KiB - all: -108.57KiB
ARM64:
HARDCODED TABLES:
BASE - 14641112
DROP DSP - 14633806 - diff: -7.13KiB
DROP MDCT - 14604812 - diff: -28.31KiB - both: -35.45KiB
DROP FFT - 14286826 - diff: -310.53KiB - all: -345.98KiB
SOFTCODED TABLES:
BASE - 13636238
DROP DSP - 13628932 - diff: -7.13KiB
DROP MDCT - 13599866 - diff: -28.38KiB - both: -35.52KiB
DROP FFT - 13542080 - diff: -56.43KiB - all: -91.95KiB
x86:
HARDCODED TABLES:
BASE - 12367336
DROP DSP - 12354698 - diff: -12.34KiB
DROP MDCT - 12331024 - diff: -23.12KiB - both: -35.46KiB
DROP FFT - 12029788 - diff: -294.18KiB - all: -329.64KiB
SOFTCODED TABLES:
BASE - 11358094
DROP DSP - 11345456 - diff: -12.34KiB
DROP MDCT - 11321742 - diff: -23.16KiB - both: -35.50KiB
DROP FFT - 11276946 - diff: -43.75KiB - all: -79.25KiB
PERFORMANCE (10min random s32le):
ARM32 - before - 39.9x - 0m15.046s
ARM32 - after - 28.2x - 0m21.525s
Speed: -30%
ARM64 - before - 36.1x - 0m16.637s
ARM64 - after - 36.0x - 0m16.727s
Speed: -0.5%
x86 - before - 184x - 0m3.277s
x86 - after - 190x - 0m3.187s
Speed: +3%
Log the warning message once per encoder instance instead.
Reviewed-by: Kieran Kunhya <kierank@obe.tv>
Reviewed-by: Michael Niedermayer <michael@niedermayer.cc>
Signed-off-by: James Almer <jamrial@gmail.com>
* commit 'cc4992aaf3dbb0af88d9727983d75636baf1f8cc':
ac3enc: allow Dolby Pro Logic IIz as the Dolby Surround EX mode.
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '4c2fd4b262347273afe97865ba451a1abde43ae6':
ac3enc: allow Dolby Pro Logic II as a preferred downmix mode.
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '9a9e2f1c8aa4539a261625145e5c1f46a8106ac2':
dsputil: Split audio operations off into a separate context
Conflicts:
configure
libavcodec/takdec.c
libavcodec/x86/Makefile
libavcodec/x86/dsputil.asm
libavcodec/x86/dsputil_init.c
libavcodec/x86/dsputil_mmx.c
libavcodec/x86/dsputil_x86.h
Merged-by: Michael Niedermayer <michaelni@gmx.at>
* commit '27631796c9d1b8146ad4a16e6539ecc08afa7565':
ac3: Only initialize float_dsp for the float encoder variant
Merged-by: Michael Niedermayer <michaelni@gmx.at>
It makes more sense for a bit mask to use an unsigned type.
The change should be source and binary compatible on all
supported systems, hence micro version bump.
Fixes a few invalid shifts.
Signed-off-by: Mans Rullgard <mans@mansr.com>
Move bit counting to the bit allocation function. Move exponent grouping to
after bit allocation. This will allow for adjustment of bandwidth parameters
during bit allocation without having to do exponent grouping multiple times.
This checks if the set of selected exponent strategies for all blocks in a
channel are in the frame exponent strategy table, and if so, writes the
table index instead of each strategy. This saves up to 7 bits per channel per
frame, so the overall effect on quality is small.
Since both the fixed-point and floating-point encoders use the FFTContext,
this no longer needs to be in a separate context. Also, when a short-transform
context is added, the same MDCT window will be used.
Andreas Rheinhardt
Andreas Rheinhardt
Lynne
Anton Khirnov
Michael Niedermayer
Tim Walker
Diego Biurrun
Vittorio Giovara
Thilo Borgmann
Justin Ruggles
Mans Rullgard
John Stebbins
Fabrice Bellard
Zdenek Kabelac